Google expands availability of its Personal Intelligence tool – News

Google LLC today significantly expanded the availability of the Personal Intelligence tool in its Gemini assistant and search engine.

The technology customizes artificial intelligence responses based on information in the user’s Google account.

Personal Intelligence made its original debut in mid-January. Initially, the service was included only in Google’s AI Pro and Ultra plans, which are priced at $19.99 and $249.99 per month, respectively. They offer more advanced AI models than free Google accounts with other benefits such as additional storage space.

Google is making Personal Intelligence available to all free users in the U.S. as part of today’s update. Consumers can access it via the Gemini app, the version of Gemini embedded in Chrome and AI Mode. The latter feature generates natural language responses to queries that users enter into Google Search.

Personal Intelligence is available on an opt-in basis. When users activate the feature, it gives Gemini access to information in more than a dozen Google services including Gmail, Drive, Maps and the company’s travel booking apps. The feature uses that data to align its responses with user preferences.

According to Google, one task that Personal Intelligence can ease is troubleshooting technical issues. If a user asks for advice on how to fix a malfunctioning printer, Gemini can sift through the user’s purchase history to find the printer’s model name. That reduces the amount of manual research involved in the process.

Trip planning is another task that Personal Intelligence promises to speed up. If a user asks about sites to see in an overseas city, Gemini can customize its response based on relevant travel bookings. For example, the AI might prioritize points of interest that are close to the hotel the user booked via Google Hotels. Furthermore, Gemini can take into account additional data points such as information about past trips.

Google also sees consumers using Personal Intelligence for a range of other tasks. The tool can recommend restaurants, help with scheduling and suggest new hobbies.

Personal Intelligence is powered by Google’s Gemini 3 series of large language models, which has a context window of 1 million tokens. The account data that the tool uses to personalize responses often exceeds that threshold. Google addressed the limitation using a machine learning method called context packaging. It enables Gemini to analyze only the data points that are most relevant to a query and thereby stay under the token limit.

Last year, Meta Platforms Inc. rolled out a similar personalization feature to its Meta AI assistant. It’s possible that other market players such as Microsoft Corp. will decide to follow suit. The software giant integrated its Copilot assistant with the consumer versions of Microsoft 365, a potentially valuable source of personalization data, last January.

Image: Google